Your Essential Guide to Palantir Foundry: Designing Decisions with Data

Lee Hyun-jong, Chief Executive Officer (CEO) of Bigster, a specialized agency, has published, Palantir Foundry: Design Your Judgments, with First Books. Drawing from his hands-on experience with Palantir’s core product, Foundry, he explores effective decision-making process design.

The author begins by addressing common organizational failures. Despite implementing Business Intelligence (BI), data warehouses, and data lakes, decision-making context is often lost as frontline staff speak in business terms, information technology (IT) teams translate to technical language, and model developers redefine in mathematical terms.

BI tools support management decisions through data analysis and visualization. Data warehouses store refined structured data for analysis, while data lakes are vast repositories holding raw data without format restrictions.

Ultimately, decisions often revert to experienced personnel’s intuition rather than relying on systems. The author tackles this issue through practical Palantir Foundry exercises, uploading simulated cosmetics manufacturing data, connecting pipelines, designing ontologies, and implementing workshop interfaces.

He candidly shares his struggles, including initial missteps with traditional extract, transform, load (ETL) approaches and difficulties translating business language into system architecture.

The author consistently poses a crucial question: Whose perspective does this data represent, and who takes responsibility when values change? While traditional platforms focus on attractive data presentation, Foundry prioritizes understanding what the numbers signify, what actions they necessitate, and who’s accountable.

A 5% defect rate becomes meaningless without connecting equipment, workers, materials, follow-up actions, and accountability.

Ontology is key. It defines field-specific concepts and relationships in computer-understandable terms. Foundry views data as real-world objects, concepts, relationships, and actions, not just tables and columns.

The book argues that fixing data in an object-relationship-state-action structure connects fragmented information into a cohesive decision-making framework, preserving processes and responsibilities within the system.

Practical explanations are detailed. Links are judgment pathways, not mere associations. The author emphasizes removing noise from workshop interfaces, focusing on immediate decisions rather than flashy visuals.

Notably, the author doesn’t present Foundry as a cure-all. He stresses that even with well-designed systems, success hinges on organizational attitudes and operational changes.

The book concludes by addressing artificial intelligence (AI) judgment challenges. While organizations often expect AI to provide optimal solutions to complex problems, the author argues that real-world operations lack isolated answers. Instead, he emphasizes defining acceptable choice ranges, boundaries, and decision-making roles in specific scenarios.

△ Palantir Foundry: Design Your Judgments / By Lee Hyun-jong / First Books / 296 pages